Research Article
Nonlinear Analysis of Electrocardiography Signals for Atrial Fibrillation
Table 1
The performances of ANN, SVM, and ELM for classifying AF groups and separating AF ECGs from normal ECGs.
| ā | Classifier | Training process time (sec) | Testing process time (sec) | Accuracy (%) |
| Classification of AF groups (N, T, S) | ANN | 44.25 | 2.27 | 94.50 | SVM | 1.75 | 0.13 | 90.15 | ELM | 0.03 | 0.005 | 96.25 |
| Separation of AF ECG from normal ECG | ANN | 34.25 | 2.15 | 97.60 | SVM | 2.30 | 0.12 | 92.35 | ELM | 0.03 | 0.005 | 99.15 |
|
|